Basement membrane-like structures containing NTH α1(IV) are formed around the endothelial cell network in a novel in vitro angiogenesis model.

Angiogenesis is a process through which new blood vessels are formed by sprouting and elongating from existing blood vessels. Several methods have been used to replicate angiogenesis in vitro, including culturing vascular endothelial cells on Matrigel and coculturing with endothelial cells and fibroblasts. However, the angiogenesis elongation process has not been completely clarified in these models. We therefore propose a new in vitro model of angiogenesis, suitable for observing vascular elongation, by seeding a spheroid cocultured from endothelial cells and fibroblasts into a culture dish. In this model, endothelial cells formed tubular networks elongated from the spheroid with a lumen structure and were connected with tight junctions. A basement membrane (BM)-like structure was observed around the tubular network, similarly to blood vessels in vivo. These results suggested that blood vessel-like structure could be reconstituted in our model. Laminin and type IV collagen, main BM components, were highly localized around the network, along with nontriple helical form of type IV collagen α1-chain [NTH α1(IV)]. In an ascorbic acid-depleted condition, laminin and NTH α1(IV) were observed around the network but not the triple-helical form of type IV collagen and the network was unstable. These results suggest that laminin and NTH α1(IV) are involved in the formation of tubular network and type IV collagen is necessary to stabilize the network.

Two trehalose-hydrolyzing enzymes from Crenarchaeon Sulfolobus acidocaldarius exhibit distinct activities and affinities toward trehalose.

Two archaeal trehalase-like genes, Saci1250 and Saci1816, belonging to glycoside hydrolase family 15 (GH15) from the acidophilic Crenarchaeon Sulfolobus acidocaldarius were expressed in Escherichia coli. The gene products showed trehalose-hydrolyzing activities, and the names SaTreH1 and SaTreH2 were assigned to Saci1816 and Saci1250 gene products, respectively. These newly identified enzymes functioned within a narrow range of acidic pH values at elevated temperatures, which is similar to the behavior of Euryarchaeota Thermoplasma trehalases. SaTreH1 displayed high KM and kcat values, whereas SaTreH2 had lower KM and kcat values despite a high degree of identity in their primary structures. A mutation analysis indicated that two glutamic acid residues in SaTreH1, E374 and E574, may be involved in trehalase catalysis because SaTreH1 E374Q and E574Q showed greatly reduced trehalose-hydrolyzing activities. Additional mutations substituting G573 and H575 residues with serine and glutamic acid residues, respectively, to mimic the TVN1315 sequence resulted in a decrease in trehalase activity and thermal stability. Taken together, the results indicated that Crenarchaea trehalases adopt active site structures that are similar to Euryarchaeota enzymes but have distinct molecular features. The identification of these trehalases could extend our understanding of the relationships between the structure and function of GH15 trehalases as well as other family enzymes and will provide insights into archaeal trehalose metabolism.

Functional dissection of the N-terminal sequence of Clostridium sp. G0005 glucoamylase: identification of components critical for folding the catalytic domain and for constructing the active site structure.

Clostridium sp. G0005 glucoamylase (CGA) is composed of a ß-sandwich domain (BD), a linker, and a catalytic domain (CD). In the present study, CGA was expressed in Escherichia coli as inclusion bodies when the N-terminal region (39 amino acid residues) of the BD was truncated. To further elucidate the role of the N-terminal region of the BD, we constructed N-terminally truncated proteins (Δ19, Δ24, Δ29, and Δ34) and assessed their solubility and activity. Although all evaluated proteins were soluble, their hydrolytic activities toward maltotriose as a substrate varied: Δ19 and Δ24 were almost as active as CGA, but the activity of Δ29 was substantially lower, and Δ34 exhibited little hydrolytic activity. Subsequent truncation analysis of the N-terminal region sequence between residues 25 and 28 revealed that truncation of less than 26 residues did not affect CGA activity, whereas truncation of 26 or more residues resulted in a substantial loss of activity. Based on further site-directed mutagenesis and N-terminal sequence analysis, we concluded that the 26XaaXaaTrp28 sequence of CGA is important in exhibiting CGA activity. These results suggest that the N-terminal region of the BD in bacterial GAs may function not only in folding the protein into the correct structure but also in constructing a competent active site for catalyzing the hydrolytic reaction.

Notch signaling pathway induces expression of type IV collagen in angiogenesis.

Mural cell adhesion is important for the localization of basement membrane components during angiogenesis, and cell-cell interactions are thought to be critical for basement membrane formation. Type IV collagen, a component of the basement membrane, and non-triple helical type IV collagen α1 chain (NTH α1(IV)) co-localize in the basement membrane of neovascular vessels. However, it remains unclear how type IV collagen and NTH α1(IV) are produced around the basement membrane. In the present study, we developed a de novo angiogenesis model using human umbilical vein endothelial cell spheroids and TIG-1 fibroblast cells and demonstrated that NTH α1(IV), probably with α1(IV) chain before forming triple helix molecule, was localized in the fibroblasts in contact with vascular endothelial cells. This localization was disrupted by DAPT, a Notch signaling inhibitor. DAPT treatment also reduced type IV collagen and NTH α1(IV) secretion in TIG-1 fibroblasts, along with diminished COL4A1 and COL4A2 gene expression. Downregulation of Notch3 in TIG-1 fibroblasts decreased the secretion of type IV collagen and NTH α1(IV). Taken together, these findings suggest that heterogeneous and homogeneous intercellular Notch signaling via Notch3 induces type IV collagen and NTH α1(IV) expression in fibroblasts and contributes to basement membrane formation in neovascular vessels.

Non-triple helical form of type IV collagen alpha1 chain suppresses vascular endothelial-cadherin mediated cell-to-cell junctions.

Non-triple helical collagen polypeptide α1(IV) (NTH α1(IV)) is a gene product of COL4A1 and is secreted as a polypeptide chain without the triple helix structure under physiological conditions. Studies have shown that NTH α1(IV) is up-regulated in and around vascular endothelial cells during neovascularization and vascular-like networks of in vitro angiogenesis models, suggesting its involvement in angiogenesis. In the present study, we examined the effect of NTH α1(IV) on endothelial cell-to-cell junctions, and we found that NTH α1(IV) suppressed VE-cadherin (vascular endothelial cadherin) mediated junctions and promoted cellular migration in human umbilical vein endothelial cell cultures. NTH α1(IV) is potentially a factor that induces VE-cadherin endocytosis and promotes neovascular sprouting and elongation. The possible mechanism entails endocytosis of NTH α1(IV) by its cellular receptor(s), Endo180 and/or other proteins, which results in the clearance of the cellular receptor(s) from the cell surface, thus inducing the endocytosis of VE-cadherin. Because the NC1 domain of the α1 chain of type IV collagen, called arresten, is considered an endogenous inhibitor of angiogenesis, it seems that the single polypeptide chain of NTH α1(IV) has conflicting functions.

Excessively activated plasminogen in human plasma cleaves VWF multimers and reduces collagen-binding activity.

Plasmin (Pm) is a serine protease that can dissolve fibrin clots. Several possible functions of Pm in blood other than fibrinolysis have been proposed. To explore the effects of Pm on primary haemostasis, we evaluated the cleavage of von Willebrand factor multimers (VWFMs) in human plasma by streptokinase (SK)-activated plasminogen (Pg) and the binding ability of the digested VWFMs to collagen. SK-activated Pg and ADAMTS13 (a VWF-cleaving enzyme) in human plasma cleaved VWFMs in conformation-dependent manners through dialysis to the urea-containing buffer. However, VWFMs in human plasma under vortex-based shear stress were cleaved by SK-activated Pg but not by ADAMTS13. These results suggested that the VWFM-cleavage sites in human plasma are exposed to some extent by vortex-based shear stress for Pm but not for ADAMTS13. Additionally, we revealed that cleavage by SK-activated Pg reduced VWFMs' binding ability to collagen, and VWFMs in human plasma were cleaved by Pm at several sites. These results suggest that SK-activated Pg degrades VWFMs, reduces their binding abilities to collagen and affects primary haemostasis. Because excessive Pg activation can degrade fibrinogen/fibrin, we propose that SK-activated Pg in blood may cause impaired primary and secondary haemostasis.

Proteolytic inactivation of ADAMTS13 by plasmin in human plasma: risk of thrombotic thrombocytopenic purpura.

Thrombotic thrombocytopenic purpura (TTP) is caused by inactivation of a von Willebrand factor (VWF)-cleaving enzyme, a disintegrin and metalloproteinase with a thrombospondin type 1 motif, member 13 (ADAMTS13), which leads to platelet-rich thrombi comprising unusually large VWF multimers. We have found that ADAMTS13 can bind to the inactivated form of plasmin. In addition, plasmin cleaves purified ADAMTS13 into several fragments and inactivates it. Hence, we hypothesized that activation of plasminogen to plasmin becomes a new-onset factor for TTP due to ADAMTS13 inactivation. Plasmin was added exogenously or activated from plasminogen by streprokinase addition in human plasma (HP). ADAMTS13 digestion and effects of the digestion on ADAMTS13 activity were evaluated. Exogenous plasmin cleaved ADAMTS13 into several fragments, but a portion of ADAMTS13 remained in full-length form. Digestion profile of ADAMTS13 with streprokinase added exogenously in HP was similar to that of ADAMTS13 with exogenous plasmin. ADAMTS13 activity measured using FRETS-VWF73 decreased to ∼40% compared with that for normal plasma. Endogenous VWF multimer-cleaving activity was attenuated more severely (∼10%). These data suggest that endogenous plasmin cleaves ADAMTS13 into fragments and reduces its activity to ∼10%. We suggest that endogenous plasmin activation alone is not sufficient to cause TTP, but plasmin activation with ADAMTS13 deficiency might increase the risk of TTP onset.

A large deletion of the PROS1 gene in a deep vein thrombosis patient with protein S deficiency.

Inherited deficiency of protein S encoded by the PROS1 gene constitutes an important risk factor for deep vein thrombosis (DVT). Nevertheless, although more than 200 deleterious genetic variations in PROS1 have been identified, causative point mutations of PROS1 gene are not detected in approximately half of protein S-deficient families. The present study investigated whether there may exist a large deletion in PROS1 that constitutes a genetic risk factor for Japanese DVT patients. A multiplex ligation-dependent probe amplification analysis was employed to identify the deletions in PROS1 in 163 Japanese patients with DVT. A large gene deletion was identified in one patient who showed 16% protein S activity and did not carry point mutations in PROS1 by DNA sequencing and it was validated by the quantitative PCR method. The deletion spanned at least the whole PROS1 gene (107 kb) and at most from the centromere located downstream of PROS1, to before the D3S3619 marker, the first heterozygous marker in the upstream of PROS1 in chromosome 3. In conclusion, a large deletion in PROS1 was shown to partly account for DVT with protein S deficiency. Screening for large deletions in PROS1 might be warranted in PROS1 causative point mutation-negative DVT patients with protein S deficiency.

Grainyhead-like 3, a transcription factor identified in a microarray screen, promotes the specification of the superficial layer of the embryonic epidermis.

The Xenopus ectoderm consists of two populations of cells, superficial polarised epithelial cells and deep, non-epithelial cells. These two cell types differ in their developmental fate. In the neural ectoderm, primary neurons are derived only from the deep cells. In the epidermal ectoderm, superficial cells express high levels of differentiation markers, while most of the deep cells do not differentiate until later when they produce the stratified adult epidermis. However, few molecular differences are known between the deep and superficial cells. Here, we have undertaken a systematic approach to identify genes that show layer-restricted expression by microarray analysis of deep and superficial cells at the gastrula stage, followed by wholemount in situ hybridisation. We have identified 32 differentially expressed genes, of which 26 show higher expression in the superficial layer and 6 in the deep layer and describe their expression at the gastrula and neurula stage. One of the identified genes is the transcription factor Grhl3, which we found to be expressed in the superficial layer of the gastrula ectoderm and the neurula epidermis. By using markers identified in this work, we show that Grlh3 promotes superficial gene expression in the deep layer of the epidermis. Concomitantly, deep layer specific genes are switched off, showing that Grlh3 can promote deep cells to take on a superficial cell identity in the embryonic epidermis.

Microarray-based identification of VegT targets in Xenopus.

The Xenopus T box family member VegT is expressed maternally in the vegetal hemisphere of the embryo. Mis-expression of VegT in prospective ectodermal tissue causes ectopic activation of mesodermal and endodermal markers, and ablation of VegT transcripts prevents proper formation of the mesendoderm, with the entire embryo developing as epidermis. These observations define VegT as a key initiator of mesendodermal development in the Xenopus embryo, and in an effort to understand how it exerts its effects we have used microarray analysis to compare gene expression in control animal caps with that in ectodermal tissue expressing an activated form of VegT. This procedure allowed the identification of 99 potential VegT targets, and we went on to study the expression patterns of these genes and then to ask, for those that are expressed in mesoderm or endoderm, which are direct targets of VegT. The putative regulatory regions of the resulting 14 genes were examined for T domain binding sites, and we also asked whether their expression is down-regulated in embryos in which VegT RNA is ablated. Finally, the functions of these genes were assayed by both over-expression and by use of antisense morpholino oligonucleotides. Our results provide new insights into the function of VegT during early Xenopus development.